Once my LX200 was  permanently set up in a roll off roof observatory, I decided to 'get into' ccd imaging. I first used  a ccd camera in September 1997. Locating targets was fairly straight forward using the large chip in the ST7.  But, after only a short time I had realised that focusing was one most difficult parts of ccd imaging. My initial set up was using the Meade electric focuser on my LX200 and operating this via the control panel in Epoch 2000. The response to the mouse button clicks was hopeless and switching between CCDOPS and Epoch was wearing out <alt> <tab> !
A separate control method other than using the PC was needed.
My second attempt was to wire a switched 12vdc supply directly to the Meade focuser, completely eliminating the PC.. This worked and I managed to obtain better results. But there was room for improvement. Lots of room !

The main reasons for working on this project were:-

1. I found achieving accurate focus was rather hit and miss. I needed a position reading to know where best focus was.
2. I use the star peak brightness value to focus, and have to go past best focus at least twice to get an idea where best focus is. The whole process of focusing needed to be faster.
3. It is worth checking focus after a period of time or when moving to a new target.
4. I was spending as much time on setting and checking focus as I was on imaging. The focus seemed to drift off quite quickly.
5. A large focus window which is slow to update was needed to keep a star in view because of the image shift the is inherent with a SCT. Image shift had to be reduced to a minimum.
6.I operate my telescope and camera remotely so additional focusing aids could not be used.
7. Precise focusing has to be obtained before optimum results can be obtained from a ccd camera.

From left to right:-
Hand controller printed circuit with digital position counter display ready for its' case.
Stepper motor drive pcb.
Focus mount assembly,  stepping motor and 25:1 gearbox.
 
 

The Mechanics'

The focus head consists of  a fixed plate and a shaft assembly housing. The two aluminum parts are connected together with four threaded shafts.
The four shafts are supported on ball races and are driven by a stepping motor and a reduction gearbox. The four shafts are synchronized by means of a timing belt. When the shafts turn, the camera is moved away or brought towards  the telescope by a controlled amount.
One step of the motor gives  1.25mm pitch / (48*25) = 0.001mm which provides plenty of resolution.
The approximate depth of focus @ f6.3 is around 0.05mm which equates to 50 steps of the motor. This means the threaded shafts need to be positioned to about 1/25 of a turn. Simple measurements have shown this is easily reached and is repeatable.

The Electronics'

The hand controller is based on a micro-controller integrated circuit. A PIC 16C84 was chosen as it has a total of 12 individually programmable input or output lines, a clock timer and software that can be written in assembler.
The counter display is a simple up-down counter module.
Four push buttons provide focus in / out, focus speed and position count reset.
Three focus speeds can be selected. Speeds are  200, 100 and 50 counts per second
The controller generates the necessary clock and direction signals to the stepping motor drive circuit. All signals are 12v.

The stepping motor drive board is based on the SAA1027 driver ic. The stepping motor connects to this, and then connects to the hand controller via a multi-core cable that runs from the observatory to the house.
 
 
Focusing with the new head.

When changing focus I can use the smallest focus box that is possible using CCDOPS which results in very fast screen updates. The star position shifts in a very small circular path about 10 pixels  (binned * 2) . Good compared to the 100 pixels shift I was getting !   I think the image movement is due to the threads on the shafts not running exactly concentric.
To set focus, a star is centered, exposure time set and the focus head adjusted.
As focus approaches,  the peak brightness value no longer increases. I make a note of the focus position. Continuing to focus in the same direction I note the position when the peak brightness value starts to drop. The focus position is then set in the center of these two readings.
There is about 150 counts of backlash in the gearbox, which has to be taken into account when changing focus direction. Although I always focus to the final position by adjusting the focus head in the same direction, moving the camera towards the telescope.
I have noticed that by not having to move the primary mirror focus adjustment, once I have focus, the telescope tends to remain focused for longer periods of time.
 
I am very please with the results so far. With more use and using new camera software that's now available which gives better information on the focused stars' profile I am hoping to improve my ccd imaging results.